Note that the tree drawn from the 
-globin amino acid sequences
corresponds with our intuition (and good external evidence) about the
evolutionary relationships among the taxa that donated those sequences. It
is by no means always the case that ``gene trees" and ``species trees"
have such a nice correspondence(Takahata[12]) (though that is
often the unspoken hope of molecular systematists). As an example of how
gene and species trees can fail to agree, consider the following
situation:
The common ancestor (O) of three species A, B, and C was polymorphic for
the gene being studied, with three alleles y, x, and 
, where
x and 
are very similar, and y different from both. The tree
which we are able to infer from present-day data on this gene may be quite
misleading, depending on which of O's alleles were passed on to its
descendants. This is likely to be a problem if one tries to use
slowly-evolving molecules with recently-diverged species, as in the tree
above
. This leads us to the problem of choosing genes.